Mass transfer lectures

Chemical Engineering
Department
SPRING 2013
Dr. Mohamed A. El-Behlil
 Molecular Mass Transfer
• Molecular diffusion
• Mass transfer law components:
– Molecular concentration:
A
nA p A
cA   
MA V RT
– Mole fraction:
cA (liquids,solids) , cA (gases)
xA  yA 
c c
 p A RT p A
For gases, y A  
P RT P
n n
– Velocity:  v  i i i vi
mass average velocity, v  i 1n  i 1

 i
i 1
n

c v i i

molar average velocity, V  i 1

c

velocity of a particular species relative to mass/molar average is

the diffusion velocity.
mol
– Flux:
A vector quantity denoting amount of a particular species that
passes per given time through a unit area normal to the vector,
given by Fick’s First Law, for basic molecular diffusion

J A  DAB c A
or, in the z-direction,
dc A
J A, z  DAB
dz
For a general relation in a non-isothermal, isobaric system,
dy A
J A, z  cDAB
dz
– Since mass is transferred by two means:
• concentration differences
• and convection differences from density differences
• For binary system with constant Vz,
J A, z c A (v A, z  Vz )
• Thus,
dy A
J A, z c A (v A, z  Vz )  cDAB
dz
• Rearranging to
dy A
c Av A, z  cDAB  c AVz
dz
• As the total velocity,
1
Vz  (c A v A , z  c B v B , z )
c
• Or

c AVz  y A (c Av A, z  cB vB , z )

• Which substituted, becomes

dy A
c Av A, z  cDAB  y A (c A v A , z  c B v B , z )
dz
• Defining molar flux, N as flux relative to a fixed z,
N A c A v A
• And finally,
dy A
N A, z  cDAB  y A ( N A, z  N B , z )
dz
• Or generalized,

N A  cDAB y A  y A (N A  N B )
• Related molecular mass transfer
– Defined in terms of chemical potential:
d c DAB d c
v A, z  Vz u A 
dz RT dz
– Nernst-Einstein relation

DAB d c
J A, z c A (v A, z  Vz )  c A
RT dz
 Diffusion Coefficient
• Fick’s law proportionality/constant
 J A, z M 1 L 2
DAB  ( 2 )( 3
)
dc A dz L t M L 1 L t
• Similar to kinematic viscosity, , and
thermal diffusivity, 
• Gas mass diffusivity
– Based on Kinetic Gas Theory
1
DAA*  u
3
–  = mean free path length, u = mean speed
2T 3 / 2  3N
DAA*  ( )1/ 2
3  P MA
3/ 2 2
A
– Hirschfelder’s equation:
1/ 2
 1 1 
3/ 2
0.001858T   
 M A M B
DAB 
P AB  D
2
– Lennard-Jones parameters  and  from tables,
or from empirical relations
– for binary systems, (non-polar,non-reacting)
 A  B
 AB   AB   A B
2

– Extrapolation of diffusivity up to 25
atmospheres
3/ 2
 P1   T2   D T1
DABT2 ,P2 DABT1 ,P1    
 P2   T1   D T2
Binary gas-phase Lennard-Jones
“collisional integral”
– With no reliable  or , we can use the Fuller
correlation,  
1/ 2
3 1.75 1 1
10 T   
DAB   MA MB 

P  v A   v B
1/ 3 1/ 3
 2

– For binary gas with polar compounds, we

calculate  by
0.196 2
  D0  *
AB
T
 where
1. 94 10 P
3 2
 A B  ,  
1/ 2
 AB
VbTb
T T /  AB
*

1/ 2
 AB   A  B 
 
   
 /  1.181  1.3 Tb
2

A C E G
 D0    
T 
* B * * *
exp(DT ) exp(FT ) exp(HT )
and 1/ 3
 1.585Vb 
 AB  A B  1/ 2
  2 
 1  1.3 

– For gas mixtures with several components,

1
D1 mixture  '
y2 / D1 2  y3' / D1 3  ...  yn' / D1 n
– with
' y2
y 
2
y2  y3  ...  yn
2
• Liquid mass diffusivity
– No rigorous theories
– Diffusion as molecules or ions
– Eyring theory
– Hydrodynamic theory
• Stokes-Einstein equation
T
DAB 
6r B
– Equating both theories, we get Wilke-Chang
eq.
DAB  B 7.4 10  B M B 
8 1/ 2

T VA0.6
– For infinite dilution of non-electrolytes in
water, W-C is simplified to Hayduk-Laudie eq.
5  1.14  0.589
DAB 13.26 10  B V A

– Scheibel’s equation eliminates B,

DAB  B K
 1/ 3
T VA
  3V  2 / 3 
8
K (8.2 10 ) 1   B  
  VA  
– As diffusivity changes with temperature,
extrapolation of DAB is by
n
( DABT1 )  Tc  T2 
 
( DABT2 )  Tc  T1 
– For diffusion of univalent salt in dilute solution,
we use the Nernst equation
2 RT
DAB 
(1 /   1 /  ) F
0 0
• Pore diffusivity
– Diffusion of molecules within pores of porous
solids
– Knudsen diffusion for gases in cylindrical pores
• Pore diameter smaller than mean free path, and
density of gas is low

• Knudsen number Kn 
d pore
• From Kinetic Theory of Gases,

u 8NT
DAA*  
3 3 M A
 • But if Kn >1, then

d pore d pore 8NT T

DKA  u 4850d pore
3 3 M A MA
• If both Knudsen and molecular diffusion exist, then
1 1  y A 1
 
DAe DAB DKA
• with
NB
 1 
NA
• For non-cylindrical pores, we estimate

D  DAe
'
Ae
2
Example 6
Types of porous diffusion. Shaded areas represent nonporous solids
– Hindered diffusion for solute in solvent-filled
pores
• A general model is

DAe D F ( ) F2 ( )
o
AB 1

• F1 and F2 are correction factors, function of pore

diameter, d
 s
d pore
• F1 is the stearic partition coefficient
 (d pore  d s ) 2
F1 ( )  (1   ) 2

 d pore
2
• F2 is the hydrodynamic hindrance factor, one
equation is by Renkin,

F2 ( ) 1  2.104  2.09 3  0.95 5

Example 7
 Convective Mass Transfer
• Mass transfer between moving fluid with
surface or another fluid
• Forced convection
• Free/natural convection
• Rate equation analogy to Newton’s cooling
equation
N A kc c A
Example 8
 Differential Equations
• Conservation of mass in a control volume:

  v n dA   dV 0
c.s. t c.v.

• Or,
in – out + accumulation – reaction = 0
• For in – out,
– in x-dir, n A, x yz x  x  n A, x yz x

– in y-dir, n A, y xz y  y  n A, y xz y

– in z-dir, n A, z xy z  z  n A, z xy z

• For accumulation,
 A
xyz
t
 • For reaction at rate rA,
rA xyz

• Summing the terms and divide by xyz,

nA, x x x  nA, x x nA, y y y  nA, y y nA, z z z  nA, z z  A
    rA 0
x y z t

– with control volume approaching 0,

    A
n A, x  n A, y  n A, z   rA 0
x y z t
• We have the continuity equation for
component A, written as general form:
 A
n A   rA 0
t
• For binary system,
  A   B 
n A  n B    rA  rB  0
t

• but n A  n B   A v A   B v B  v
• and rA  rB
• So by conservation of mass,

v  0
t
• Written as substantial derivative,
D
 v 0
Dt
– For species A,
D A
 j A  rA 0
Dt
• In molar terms,
c A
N A   RA 0
t
– For the mixture,
c A  cB 
N A  N B   ( RA  RB ) 0
t
– And for stoichiometric reaction,
c
cV   ( RA  RB ) 0
t